Imputation of Public Opinion Data with Large Language Models
This work presents the first evaluation of LLM imputation against proper multiple-imputation metrics and statistical baselines on more than 5 million imputed responses from four public opinion surveys under various missingness mechanisms and highlights the importance of fit-for-purpose evaluation for LLM imputation.
Tobias Holtdirk, Georg Ahnert, J. Sakshaug et al.
· 2 citations